1

Culture and Cognition

Daniel M.T. Fessler (Center for Behavior, Evolution, & Culture and Department of

Anthropology, UCLA)

Edouard Machery (Department of History and Philosophy of Science, University of

Pittsburgh)i

To appear in The Oxford Handbook of Philosophy of Cognitive Science, E. Margolis, R.

Samuels, and S. Stich, eds. Oxford University Press

Humans are unique among animals for both the diverse complexity of our cognition and

our reliance on culture, the socially-transmitted representations and practices that shape

experience and behavior. Adopting an evolutionary psychological approach, in this essay

we consider four different facets of the relationship between cognition and culture. We

begin with a discussion of two well-established research traditions, the investigation of

features of mind that are universal despite cultural diversity, and the examination of

features of mind that vary across cultures. We then turn to two topics that have only

recently begun to receive attention, the cognitive mechanisms that underlie the

acquisition of cultural information, and the effects of features of cognition on culture.

Throughout, our goal is not to provide comprehensive reviews so much as to frame these

issues in such a way as to spur further research.

2

1 Psychological universals

Psychological universals can be defined as those traits, processes, dispositions, or

functions that recur across cultures, with at least a subset of each population (e.g.,

individuals of a specific gender or at a specific developmental stage) exhibiting the trait.

The search for psychological universals has a long tradition, as illustrated by Darwin’s

(1872) investigation of universal emotions, behaviorists’ search for universal laws of

learning (Hull, 1943), and the Chomskian approach to language and cognition (e.g.,

Chomsky, 1986). This tradition has been in part motivated by the desire to establish the

“psychic unity” of humanity.

Because traits may recur across cultures due to cultural influences alone (via

common cultural descent, cultural diffusion, or cultural evolutionary convergence), the

strongest test of the universality of a given psychological trait is to search for it across

maximally disparate cultures. One methodological concern, however, is that whether or

not a trait is identified in different cultures will depend in part on how the trait is defined.

For instance, if the specific circumstances that trigger shame in the USA are included in

the definition of this emotion, shame is unlikely to qualify as a psychological universal;

however, if the eliciting conditions are described more abstractly, shame is a good

candidate for a universal emotion (e.g., Fessler, 2004). It is worth keeping this in mind to

avoid empty verbal controversies about the universality of traits (Mallon & Stich, 2000).

Relatedly, for many traits, similarities and differences across cultures will coexist

(Brown, 1991; Norenzayan & Heine, 2005). Thus, shame might be present in all cultures,

yet be triggered by different circumstances, or be expressed differently (see the notion of

culture-specific emotion display rules in Ekman & Friesen, 1971).

3

1.1 Generatively entrenched homologies

Because cultures vary tremendously with respect to their ecology, social organization,

scale, and technology, and because cultural variables affect cognitive development, one

might wonder why psychological universals exist at all. The answer is simple for those

traits the development or acquisition of which reflects universal properties of physical or

social environments, e.g., the belief that water is wet, or the distinction between males

and females. The answer is not so straightforward for other psychological universals,

because cultural variables could plausibly affect their development. One might argue that

such universals are the product of evolution by natural selection, and that natural

selection tends to select for species-typical traits (Tooby & Cosmides, 1992). However,

this would be a mistake on two counts. First, much recent research emphasizes that

natural selection has favored particular forms of phenotypic plasticity in humans,

including the capacity to adapt to, and exploit, parochial cultural information. Second,

one cannot presume that natural selection generates homogeneity. In most species, many

traits are adaptive polymorphisms, either as a result of frequency-dependent selection or

as an adaptive response to environmental variation in the species’ range.

So, where do psychological universals come from? Some traits may be

psychological universals because they are homologies—features possessed by humans

and their relatives by virtue of common descent—that are generatively entrenched. A trait

is generatively entrenched if its development is a necessary condition for the

development of other traits (Wimsatt, 1986). Most modifications of a generatively

entrenched trait are selected against because they prevent the development of these other

4

traits. If a psychological trait in humans is homologous to traits in other species, then,

given the general absence of culture outside of our own lineage, it follows that the trait

originally evolved in a species that had little (if any) capacity for culture. If this trait

became generatively entrenched, then natural selection had little scope to act on its

development, which remained insensitive to cultural variables.

The approximate numerical sense provides a clear example of a generatively

entrenched homology. Research in the USA and in Europe has established that children

and adults possess an approximate number sense (Hauser & Spelke, 2004; Piazza &

Dehaene, 2004), being able to approximate the cardinality of sets of visually presented

objects or of sequences of sounds without counting, to compare the cardinality of

different sets or sequences, and to approximate the results of adding several sets of

objects. The accuracy of people’s numerical evaluation obeys Weber’s law: the mean

evaluation is identical to the cardinality of the target set, and the evaluation’s standard

deviation linearly increases as a function of the cardinality of the target set. Because

evaluation is thus increasingly noisy, the accuracy of the numerical comparison between

sets or sequences increases as a function of the distance between the cardinality of the

sets or sequences to be compared. These are the signature properties of an analogical

encoding of the cardinality of sets or sequences.

Recent studies by Pica, Lemer, Izard, and Dehaene (2004) and Gordon (2004)

provide strong evidence that this approximate number sense is a psychological universal.ii

Pica et al. studied approximate estimation, comparison, and addition among the

Mundurukú, a small-scale society in Brazilian Amazonia having limited contact with

non-indigenous people. Most Mundurukú have not received any formal education. Their

5

language has words for only the numbers 1 to 5; above 5, the Mundurukú rely on

locutions signifying “some” or “many.” Strikingly, the numbers 3, 4, and 5 are also used

to refer to approximate quantities. For instance, in Pica et al.’s experiments, the word for

4 was used for sets of 4 and 5 objects, and the word for 5 was used for sets of 5 to 9

objects. In spite of the differences between the Mundurukú counting system and that in

European (and other) languages, and in spite of the many other differences between the

respective cultures, the Mundurukú’s approximate number sense is identical to

Europeans’. Their performances in estimation, comparison, and addition tasks show the

signature properties of the analogical system assumed to encode the cardinality of sets or

of sequences. Gordon (2004) found similar results with the hunter-gatherer Pirahã tribe in

the Lowland Amazonia region of Brazil. Most strikingly, the language spoken by the

Pirahã has words for only 1, 2, and 3. Nonetheless, their performances in tasks tapping

into their approximate number sense were very close to Europeans’ and Americans’,

providing further evidence for the universality of the approximate number sense.

The approximate number sense, evident in cultures as diverse as small-scale

hunter-horticulturalist societies and modern, technologically complex societies, is also

present in numerous animal species (Hauser & Spelke, 2004). Thus, in line with our

discussion of the origins of psychological universals, the approximate number sense is

plausibly a generatively entrenched homology.

1.2 Canalized traits

Not all psychological universals are generatively entrenched homologies. A number of

uniquely human psychological traits are also universal because their development has

6

been canalized during the evolution of human cognition. Natural selection selects against

development pathways that rely on specific environmental inputs when these

environmental inputs vary, when variation in these environmental inputs cause the

development of variable traits, and when there is a single optimally adaptive variant

(Waddington, 1940).iii When this happens, natural selection buffers the development of

the relevant traits against environmental variation by selecting for developmental

pathways that do not depend on these environmental inputs. This phenomenon, known as

canalization, likely explains the origins of some psychological universals. Note that, in

contrast to the explanation of the evolution of psychological universals examined above,

this second account can explain the universality of psychological traits that are not

homologies.

Research on so-called folk theories provides some of the best evidence for such

universals (Sperber & Hirschfeld, 2004; Boyer & Barrett, 2005). Folk theories are

domain-specific, often implicit bodies of information that people use to reason. Although

many folk theories vary across cultures, in some domains, folk theories have a universal

core; folk biology and folk psychology are two such cases.

Some aspects of folk biology vary across cultures: in some cultures people have

much more extensive biological knowledge than in others, and some reasoning strategies

about the biological domain are found only in some cultures (for review, see Medin,

Unsworth, & Hirschfeld, 2007). Despite such heterogeneity, across cultures, people

classify animals and plants in a similar way (Berlin, Breedlove, & Raven, 1973; Berlin,

1992; Atran, 1990, 1998; Medin & Atran, 2004; Medin et al. 2007). Plants and animals

are organized into hierarchically organized taxonomies of kinds that include (at least)

7

three levels: a “generic species” category (e.g., dogs and cedars), a superordinate

category of biological domains (e.g., animals and plants), and a subordinate category of

species varieties (e.g., particular breeds or strains). At any level, membership in a kind is

exclusive. For instance, no animal is both a dog and a cat or a fish and a mammal.

From a cognitive perspective, the generic species level is of particular importance.

Atran and colleagues have shown that while Itza’ Maya’s biological knowledge is much

more extensive than American undergraduates’, both Itza’ Maya and American

undergraduates avoid generalizing biological properties to the members of categories

whose level is above the generic-species level (Coley, Medin, & Atran, 1997).

Furthermore, in a range of diverse cultures, membership in generic species is associated

with “psychological essentialism” (Medin & Ortony, 1989; Gelman, 2003; Medin &

Atran, 2004): people believe that membership in a biological kind is associated with the

possession of a causal essence—that is, some property or set of properties that define

membership in the kind and cause the members of this kind to possess the kind-typical

properties independently of their rearing environment. An essentialist disposition has

been found among American children and adults (Keil, 1989; Gelman & Wellman,

1991), Yucatek adults (Atran, Medin, Lynch, Vapnarsky, et al., 2001), Brazilian adults

(Sousa, Atran, & Medin, 2002), and among children and adults from Madagascar (Astuti,

Solomon, & Carey, 2004)

Similarly, in spite of many differences in the way people across cultures explain

their and others’ behavior (Lillard, 1998), psychologists have identified a universal core

in folk psychology. In every known culture, people explain behavior in mentalistic terms,

i.e., by ascribing mental states such as beliefs and desires (Wierzbicka, 1992).

8

Furthermore, in their meta-analysis of children’s performances in the false belief task,

Wellman, Cross, and Watson (2001) have shown that children’s understanding of beliefs

develops similarly across cultures. They report studies conducted in Western cultures

(USA, Austria, Australia), Eastern cultures (Japan, Korea), among the hunter-gatherer

Baka (Avis & Harris, 1991), and among Quechua-speaking Peruvian Indians (Vinden,

1996). Although culture affects how early children come to understand that beliefs can be

false, 3-year-old children from Western, Eastern, African, schooled, and non-schooled

cultures initially fail to understand this, then progressively come to grasp the notion of

false beliefs.

2 Cross-cultural psychological diversity

An entirely different approach to the relations between culture and cognition focuses on

differences across cultures. Many differences are best described as ethnographic rather

than psychological per se: because across cultures, people live in different social and

physical environments, and different cultural framings thereof, and have correspondingly

different experiences, their beliefs, concepts, and desires—in brief, the contents of their

minds—will often similarly vary. However, looking beyond such differences, scholars

have explored the effects of cultural variation on cognitive processes, personality, and

perception (for an extensive review, see Cohen & Kitayama, 2007). The search for cross-

cultural differences is deemed successful to the extent that the differences across cultures

are marked and are explained by relevant differences among these cultures.

The search for cross-cultural psychological differences has a long history.

Particularly, numerous scholars have addressed the role of linguistic differences in

9

producing psychological differences. The anthropologist Edward Sapir and the linguist

Benjamin Whorf famously proposed that the syntax and the vocabulary of different

languages promote irreducibly different patterns of thought—what is known as the Sapir-

Whorf hypothesis (Whorf, 1956). Similarly, Soviet psychologists, particularly Lev

Vygotsky and Alexander Luria, argued that languages as well as social activities (e.g.,

counting routines) constitute tools that allow children to develop symbolic thinking

(Vygotsky, 1986).

2.1 Proximal origins of cross-cultural differences: Extended cognition

Traditionally, psychologists and anthropologists searching for cross-cultural differences

have given little thought to the evolutionary origins of this diversity, assuming that

evolutionary considerations were only relevant for universal traits, or that evolution was

only relevant in so far as it produced an undifferentiated “capacity for culture.” As

discussed in the next two sections, recent theories and findings belie these assumptions.

Here, we successively focus on two proximal causes of cross-cultural psychological

diversity.

First, while cognitive science has tended to be methodologically solipsist (Fodor,

1980), neglecting the social and physical environment in which cognition takes place, an

influential approach, termed “extended cognition,” now insists that social practices, such

as counting routines and formal education, as well as physical artifacts affect (or, in some

formulations of this idea, are constitutive of) people’s cognitive processes (e.g., Hutchins,

1995; Clark, 1997). Because practices and artifacts vary tremendously across cultures,

their effect on the mind is a potent source of cross-cultural diversity.

10

Language is one of the social practices that can potentially cause cognition to

differ substantially across cultures. Since the 1990’s, a flurry of cross-cultural work in

linguistics and psychology has revived interest in the Sapir-Whorf hypothesis (for

reviews, see Gentner & Goldin-Meadow, 2003; Gleitman & Papafragou, 2005; Chiu,

Leung, & Kwan, 2007). We consider in turn the research on spatial orientation and the

more decisive research on color perception.

Levinson and colleagues have shown that languages encode spatial orientation in

a variety of ways (Levinson, 2003; Pederson, Danzinger, Wilkins, Levinson, et al., 1998).

They identify three main ways of describing the location of objects. Speakers who use an

intrinsic frame of reference locate objects by describing the relations between these

objects (the spoon is besides the plate). Speakers who use a relative frame of reference

locate objects by describing the position of these objects in relation to themselves and

others (the knife is on my/your right). Speakers who use an absolute frame of reference

locate objects by using cardinal directions (the knife is west of the plate).

Levinson and colleagues have argued that these linguistic differences affect

people’s spatial reasoning. In the rotation experiment (Pederson et al. 1998), subjects are

shown an array of objects displayed on a table in front of them. Then, they are asked to

turn by 180o. They are then given the objects and asked to recreate the original array on a

new table. If speakers of a language with a predominantly relative frame reference also

reason relatively, they should preserve the orientation of the objects with respect to their

own body: the object that was on the subject’s left on the first table should be on her left

on the second table, and so on. If speakers of a language with a predominantly absolute

frame reference reason absolutely, they should preserve the absolute orientation of

11

objects, thereby changing the orientation of the objects with respect to their own body:

the object that was on the subject’s left on the first table should now be on her right on

the second table. As predicted, Dutch and Japanese subjects, whose languages use a

relative frame of reference, preserved the relative orientation of the objects, while

Mayans, whose language uses an absolute frame of reference, preserved the absolute

orientation of objects. Levinson and colleagues take this and other findings to support the

Sapir-Whorf hypothesis.

This finding has, however, been criticized (Gallistel, 2002; Li & Gleitman, 2002).

Li and Gleitman (2002) have shown that when a salient object is present in their physical

environments, American subjects can be primed to replicate the first array in an absolute

manner. Since subjects who replicated the array of objects in a relative manner and those

who replicated it in an absolute manner speak the same language (English), it would

seem that the linguistic differences between Teztlan and Dutch (or Japanese) do not

explain Pederson et al.’s original findings (but see Levinson, Kita, Haun, & Rasch, 2002).

Li and Gleitman propose that the task description of the rotation experiment, “Make it the

same,” is ambiguous, because there are two different ways to reproduce the original array

of objects. Subjects use the fact that their language relies predominantly on a relative or

on absolute frame of reference to disambiguate the task. Thus, Levinson and colleagues’

work does not show an effect of language on thought, but rather an effect of language on

the interpretation of a linguistic expression. Furthermore, Li, Abarbanell, and Papafragou

(2005) have shown that Teztlan speakers are not only able to use a relative frame of

reference to solve spatial problems, but are better at doing it than at using an absolute

frame of reference—in clear contrast to the Sapir-Whorf hypothesis.

12

This is not to say that the Sapir-Whorf hypothesis is unsupported by recent

research, as shown by the research on color. The color lexicon varies tremendously

across languages (Berlin & Kay, 1969). While English has 11 basic color terms, the Dani,

a hunter-horticulturalist society in Papua New Guinea, use only two, one for light colors

and one for dark colors. Color vocabulary is not entirely arbitrary, however (Kay &

Regier, 2006). Focal colors constrain languages’ color vocabulary: Regier, Kay, and

Cook (2005) found that the best examples of color terms for 110 languages from non-

industrialized societies cluster around the focal colors. Furthermore, colors tend to be

grouped by similarity. Finally, Kay and Regier (2007) have shown that the boundaries

between color terms are not arbitrary, but rather map closely across languages.

Because color vocabulary varies across languages, one might wonder whether

people’s perception and memory of colors vary across linguistic communities. Heider

(1972) and Heider and Olivier (1972) have answered negatively to this question, as the

Dani’s limited color vocabulary seems to have limited effect on their color perception

and memory. Heider and Olivier showed Dani and American subjects a color patch. After

a 30-second interval, subjects were shown an array of similar color patches and were

asked to identify the original patch in this array. The pattern of color recognition was

very similar between the two groups. Particularly, in both groups, focal colors were

recognized more easily (that is, less confused with other colors) than non-focal colors.

This body of evidence against the Sapir-Whorf hypothesis has been challenged in

recent years (Davidoff, Davies, & Roberson, 1999; Roberson, Davies, & Davidoff, 2000;

Roberson, Davidoff, Davies, & Shapiro, 2005; see also Lucy & Schweder, 1979; Kay &

Kempton, 1984). Davidoff and colleagues (1999; Roberson et al., 2000) have focused on

13

color perception among the Berinmo in Papua New Guinea, who have only five basic

color terms. They failed to replicate Heider and Olivier’s (1972) experiments. In contrast

to Americans, the Berinmo were not more accurate at recognizing focal colors than non-

focal colors. Furthermore, when Davidoff and colleagues compared the Berinmo’s

pattern of color confusion with their pattern of color naming and with the American

pattern of color confusion, they found that the Berinmo pattern of color confusion was

more similar to the Berinmo pattern of color naming than to the American pattern of

color confusion. In line with the Sapir-Whorf hypothesis, this suggests that the Berinmo’s

color vocabulary affects their color memory.

Additional evidence shows that the effect of color terms on cognition is, at least

sometimes, driven by the on-line use of color words (Roberson & Davidoff, 2000;

Winawer, Witthof, Frank, Wu, et al., 2007). Similarity judgments and color identification

(identifying two patches as being the same color) are affected by the boundaries between

color categories drawn by the color terms in the subjects’ languages. Winawer et al.

(2007) simultaneously presented subjects with a target color patch on the one hand and a

pair of color patches on the other. Subjects were asked to determine which of the two

patches in the pair was identical to the target patch. Reaction times show that this color

identification task was easier when two different color terms (in the subject’s color

vocabulary) applied to the two patches in the pair than when the same term applied to

both patches—a form of categorical perception. Importantly, this and other effects of

color vocabulary disappear when a verbal dual task prevents subjects from articulating

subvocally.

14

2.2 Proximal origins of cross-cultural differences: Effects of environmental differences

A second approach to the proximal origins of cross-cultural differences notes that people

typically have various processes and strategies for fulfilling a given psychological

function (for instance, categorizing, reasoning inductively, making decisions under

uncertainty, etc.). In any given environment, these strategies do not equally well fulfill

their functions. For instance, the different types of spatial orientation do not work equally

well in all environments. As a result, people can learn to rely on the processes and

strategies that are most efficient in their environments. It is therefore important that social

and physical environments vary across cultures. Indeed, culturally transmitted norms

directly shape people’s social environments, and cultural practices can powerfully modify

people’s physical environments. Culture is thus a source of diversity in social and

physical environments; hence, across cultures, people might come to learn to rely on

different processes and strategies, because these are the best ways to fulfill the relevant

functions in the environments they inhabit.

It is important to note that the two proximal causes we have considered lead to

two different forms of cross-cultural psychological diversity. The idea that artifacts and

social practices affect or are constitutive of people’s cognitive processes implies that

particular cognitive processes (and other psychological traits) might exist in some

cultures, but not in others. In contrast, the idea that people learn to rely on the strategies

and processes that are most efficient in their environment suggests that the same

processes (and other psychological traits) are present (if only in nascent form) in all

cultures, but are differently employed.

15

Recent work on cultural differences in attention and reasoning provides a good

example of the second type of cross-cultural psychological diversity.iv Nisbett and

colleagues distinguish two cognitive styles (Nisbett, Peng, Choi, & Norenzayan, 2001;

Nisbett, 2003; Norenzayan, Choi, & Peng, 2007). The analytic cognitive style involves

detaching focal objects from their context (field independence), focusing on the

properties of objects in contrast to relations between objects, relying on rules to classify

and reason, and appealing to causal explanation. By contrast, the holistic cognitive style

involves paying attention to the context (field dependence), focusing on the relations

between objects, and relying on similarity to classify and reason. Nisbett and colleagues

have gathered an impressive body of evidence showing that Westerners exhibit an

analytic cognitive style, while East Asians display a holistic cognitive style.

Westerners’ attention abstracts objects from their context, while East Asians’

attention relates them to their context. In the rod-and-frame test, subjects are shown a rod

inside a frame and are asked to adjust the rod to a vertical position. People are considered

field-dependent to the extent that their judgment is affected by the verticality of the

frame. Ji, Peng, and Nisbett (2000) found that Chinese subjects are more field-dependent

than American subjects. These different patterns of attention affect Westerners’ and East

Asians’ perception (for review, see Nisbett & Miyamoto, 2005). Using an eye-tracking

method, Chua, Boland, and Nisbett (2005) have shown that Chinese and American

students have different patterns of visual exploration of a scene, Americans focusing on

the main object of the scene, and Chinese paying greater attention to the background.

They propose that the differences between Westerners’ and East Asians’ attentional

16

patterns might result from the differences between the visual scenes that are characteristic

of the two cultures.

Westerners and East Asians also reason differently in a large number of contexts.

According to Norenzayan, Smith, Kim, and Nisbett (2002), when asked to assess the

similarity between a target object and members of two different categories, East Asians

rely on the family resemblance of the target object to the members of each category,

while Euro-American look for properties that are necessary and sufficient for belonging

to one of the categories. Furthermore, although East Asians are perfectly able to reason

according to the rules of propositional logic, they are less disposed than Westerners to do

so.

3 The cognitive mechanisms underlying culture acquisition

As exemplified by the work reviewed above, a substantial body of research now

documents the extent to which cultural information can play an influential, at times even

determinative, role in cognitive processes. Missing from much of this literature,

however, are ultimate explanations as to why the human mind is so reliant on, or plastic

with respect to, cultural information. Some scholars content themselves with the

generalization that because much of human social, economic, and even biological life is

structured by culture, the general propensity to think in the same manner as those around

one evolved because it facilitates coexistence. Congruent with the “extended cognition”

perspective described earlier, others argue that human cultures themselves owe their

existence to the effectiveness with which cultural concepts, ways of thinking, and

artifacts extend basic human information-processing capacities, thereby bootstrapping

17

our innate potential to a higher level of behavioral complexity. Without contesting either

of these generalizations, a nascent school of thought adopts a more explicitly mechanistic

evolutionary perspective on the relationship between culture and cognition. This

perspective begins with the long-recognized observation that, to a much greater extent

than is true of other species, humans depend on cultural information to cope with the

challenges posed by their physical and social worlds. The adaptive significance of

cultural information suggests that natural selection can be expected to have crafted the

human mind so as to maximally exploit this resource. We suggest that the evolved mental

mechanisms that serve this goal fall into two general categories, reflecting differences in

the degree of specificity of the types of information that they acquire (see also Boyer,

1998 and Henrich & McElreath, 2003 for relevant discussions). We turn first to

mechanisms dedicated to the acquisition of specific bodies of knowledge.

3.1 Domain-specific cultural information acquisition mechanisms

Two principal obstacles confronting learners who seek to benefit from others’ knowledge

are the richness of the informational environment and the incompleteness of the

discernable information therein. First, human behavior is enormously complex, varying

across contexts and persons, while linguistic utterances convey information ranging from

the trivial to the life-saving. If, as is often tacitly presumed, learners were indiscriminate

sponges, then a) learners would often fail to understand how to apply what they have

learned, and b) learners would fail to properly prioritize their acquisition efforts, often

resulting in both precocity in domains irrelevant to the learner and retardation in relevant

domains. Second, much social learning involves the problem of the poverty of the

18

stimulus, as many actions and utterances explicitly present only fractional portions of the

information that motivates them (Boyer, 1998). We suggest that, for many domains of

learning, natural selection has addressed both of these problems by endowing the mind

with inborn mechanisms, possessing considerable content, that serve to structure the

acquisition of cultural information. Such a system can address the prioritization problem

via variable motivational valence, as the attentional resources that various mechanisms

command in pandemonium-style competition can be calibrated by natural selection to

reflect the relative importance of acquiring cultural information in the respective

domains: to foreshadow the discussion that follows, acquiring cultural information about

dangerous animals likely has recurrently had a greater impact on children’s survival than

has learning to perform adult rituals, and, correspondingly, natural selection appears to

have crafted the mind such that children find the former much more interesting than the

latter. Likewise, by adjusting the developmental timing of such mechanisms, natural

selection can ensure that learners acquire the cultural information that is most relevant to

the fitness challenges characteristic of their current stage of life (for example, we

postulate that young children are more interested in dangerous animals than they are in

courtship behaviors). Finally, as has been argued extensively for the case of language

acquisition, innate content can help to overcome the poverty of the stimulus problem, as

such structure can serve as a foundation that narrows the possible referents or

implications of statements and actions.

We believe that investigations aimed at uncovering what we term domain-specific

cultural information acquisition mechanisms (DSCIAMs) can shed considerable light on

how, when, and why knowledge is acquired from others. In designing such

19

investigations, it is important to recognize that, for any given learning domain, three

factors can be expected to constitute necessary conditions for the evolution of a

DSCIAM. First, the domain must have been of substantial and relatively uniform

importance to biological fitness across the diverse socioecological circumstances that

characterized ancestral human populations, as this will have provided the steady selection

pressure necessary for the evolution of a complex adaptation. Second, the domain needs

to involve content that will have varied significantly across said circumstances as, on the

one hand, this precludes the evolution of extensive innate knowledge, and, on the other

hand, this maximally exploits culture’s ability to effectively compile information of

parochial relevance. Lastly, the domain must be one in which individual learning

through trial-and-error or direct observation would have been either very costly or

impossible much of the time under ancestral circumstances.

To illustrate the above three factors, consider Barrett’s (2005) proposal that a

dedicated mechanism or set of mechanisms facilitates learning about dangerous animals.

Dangerous animals were a persistent threat to ancestral humans. A few classes of

dangerous animals will have been ubiquitous, sharing key perceptual features across

disparate environments. For example, poisonous snakes are found in most of the

ecosystems inhabited by humans. While snakes vary in their morphological details, all

snakes share the same basic body plan. This combination of a significant and recurrent

source of selection pressure and uniform perceptual features allowed for the evolution of

a template-driven learning mechanism that requires only minimal input to produce a fear

of snakes, a homologous trait shared with other primates (Öhman & Mineka, 2003).

Notably, however, in contrast to the case of snakes, many dangerous animals are either

20

confined to discrete geographical areas, do not exhibit categorically-distinguishing

features, or both. As a consequence, natural selection cannot construct a learning

mechanism with the same type of content as that responsible for the fear of snakes—at

most, selection can assign innate salience to cues, such as large sharp teeth, that are

imperfectly associated with dangerousness. Nevertheless, because cultures can be relied

upon to contain information about the identity and attributes of, and strategies for dealing

with, locally dangerous animals, natural selection could construct a mechanism dedicated

to acquiring this kind of information from others. Importantly, fulfilling the third

criterion listed above, learning about dangerous animals from other actors is almost

always vastly cheaper than learning through trial and error.

With regard to the informational challenges, described earlier, that confront the

human learner, Barrett (2005) suggests that what we would term a dangerous-animal

DSCIAM likely contains conceptual primitives (“animal,” “dangerous”) that a) are

rapidly mapped onto local lexical terms, b) have high salience (leading to enhanced

attending to, and retention of, any co-occurring information), c) assist in solving

problems of reference and inference (e.g., the learner presumes that statements

concerning dangerousness refer to whole species rather than individual instances, that

dangerousness will loom large in others’ minds as well, allowing for inference as to the

topic being discussed, etc.) and d) structure the manner in which information is organized

and stored (e.g., the learner constructs a danger-based taxonomy of animals, etc.). This

mechanism may also be linked to other mechanisms that govern defensive strategies

(e.g., freeze, hide, flee, seek arboreal refuge, etc.) such that learning consists of

reinforcing one of a number of preexisting potential responses. Next, because the threat

21

posed by dangerous animals begins early in life (and, all else being equal, is often

inversely proportional to an individual’s size), this mechanism can be expected to begin

operating early in development. Finally, children are expected to not only preferentially

attend to and retain cultural information about dangerous animals, but, moreover, to

actively pursue such information (e.g., more frequently asking questions regarding

dangerous than non-dangerous animals, allocating time to social contexts in which

information regarding dangerous animals is likely to be available, etc.).

What other domains of learning might meet the triple criteria of universally high

fitness relevance, parochial content, and costly individual learning necessary for the

evolution of a DSCIAM? Below we list a number of possibilities:

• Diet. Humans are generalists, capable of subsisting on a wide variety of foods.

However, many ingestible items are inedible or, worse, poisonous. Obtaining nutrition

and avoiding toxins are principal determinants of fitness, yet the features of both edible

and poisonous substances differ widely across ecosystems. As in the case of predators’

sharp teeth, a few perceptual cues (e.g., sweet taste versus bitter taste) offer some

guidance in acquiring locally-useful information, but, once again, these cues are

imperfect and incomplete. Importantly, a) cultures can be reliably expected to contain

locally-useful information in this domain, and b) the costs of individual trial-and-error

learning will generally be vastly higher than the costs of cultural learning (consider the

two ways in which one might obtain the answer to the question “Is this mushroom

poisonous?”). With regard to timing, acquiring competence in this domain first becomes

important at weaning, then becomes increasingly important with greater mobility and,

eventually, foraging self-sufficiency. Lastly, it is likely that a phylogenetic precursor of a

22

dietary DSCIAM set the stage for the evolution of such a mechanism, as social learning

plays a role in the dietary behavior of a variety of language-less mammalian generalists

(Galef & Giraldeau, 2001).

• Social structure. All societies operate in part based on a set of interrelated culturally-

defined roles that demarcate identities and prescribe actions. Roles dictate social behavior

in a relational fashion, as the appropriate actions depend not simply on one’s own role,

but on the interface between that role and the role of the individual with whom one is

interacting (e.g., the appropriate actions of a “daughter” depend on with which parent she

is interacting). Accordingly, actors must acquire information concerning not only their

own roles, but also many that they will never occupy themselves. Relatedly, coalitions

and corporate groups (associations having a collective identity such that the actions of

one member affect all) are of critical importance, as a) they are often the principal

sources of aid, and b) social conflict often occurs along such lines of cleavage. Because

both coalitions and corporate groups form along culturally-defined lines, actors benefit

from acquiring role-related cultural information regarding the nature, composition, and

boundaries of such groups. Lastly, at a still larger scale, actors benefit from preferential

assortment with those who share a similar cultural background, as this facilitates

coordination and cooperation; conversely, it often pays to understand disjunctures or

conflicts between those who resemble oneself and others who possess different cultural

affiliations.

From roles all the way up to ethnic groups, the scope of social structural

knowledge to be obtained is potentially very large; importantly, it is also highly

parochial. While social structures share common conceptual elements, such as

23

distinctions based on age, gender, kinship, hierarchy, communality, and group

membership (Brown, 1991), these core elements map only indirectly onto local forms—

most roles, for example, cannot be defined in terms of these primitives. Because social

relations would have been a fundamental determinant of fitness in ancestral societies, we

can expect selection to have crafted mechanisms specific to the task of acquiring social

structural information (Hirschfeld, 1996, 2001). Such mechanisms likely contain the

aforementioned conceptual primitives, using these as scaffolds to construct simple role

concepts early in life that can, in turn, subsequently serve as the foundations for more

complex concepts. Consistent with this premise, learners seem predisposed to a) acquire

labels for, and information concerning, a variety of social categories, b) rapidly recognize

coalitions and corporate groups, and c) apply essentialist reasoning to larger categories of

individuals (technically termed ethnic groups) who resemble one another by virtue of

their common possession of indices, such as dialects and other markers of ethnicity, of

shared cultural knowledge (Hirschfeld, 1996, 2001; Gil-White, 2001; Machery &

Faucher, 2005).

• Tool Use. Skill in using tools is a determinant of fitness in all known small-scale

societies. Some of the informational basis of tool use skill can be obtained though trial-

and-error learning, as artifacts’ affordances will often bias experimental efforts in the

direction of techniques congruent with the tool’s design. However, in many instances,

trial-and-error learning will be more expensive (in terms of time, and in terms of risk of

injury to self, the tool, or the other objects or persons) than social learning; even in

technologically simple societies, acquiring mastery of some tool techniques (e.g.,

flintknapping) is laborious, or even impossible, without cultural information. We can

24

therefore expect natural selection to have crafted DSCIAMs dedicated to this domain (see

also Csibra & Gergely, 2006). These mechanisms may contain conceptual primitives,

such as “piercing tool,” “cutting tool,” “lever,” “carrying tool,” “container,” and so on,

that aid in the acquisition of cultural information linking a specific tool, a specific

objective, and a specific technique. Testifying to our reliance on socially-transmitted

information in conceptualizing artifacts, even brief exposure to another’s use of a novel

artifact suffices to create a categorical representation of its function (Defeyter & German,

2003). The predisposition to conceptualize tools in a socially-determined fashion is

apparently so strong that, even in technologically simpler societies, such categorization

reduces the improvisational application of tools to novel purposes, a constraint termed

“functional fixedness” (German & Barrett, 2005).

The above discussion is intended to be illustrative, not exhaustive; other domains

in which DSCIAMs likely operate include navigation, fire building (Fessler, 2006),

disease avoidance, gathering, hunting (Barrett, 2005), courtship and mateship, and,

perhaps most importantly, morality (Haidt & Joseph, 2004). Moreover, research aimed at

mapping DSCIAMs will likely overlap somewhat with, and should drawn on, existing

work on psychological universals of the type discussed in Section 1. For example, the

postulated dangerous-animal DSCIAM is likely either part of, or linked to, broader

mechanisms responsible for learning about living things; these mechanisms likely contain

considerable innate content, generating the universality of core features of folk biology

described earlier. Our goal here, however, is not simply to expand the scope of the

search for psychological universals, but rather to direct attention to the means whereby

cultural information is acquired.

25

We hope that, by considering the three criteria of universal ancestral fitness

relevance, parochial content variation, and high-cost individual learning, scholars will

identify numerous areas in which DSCIAMs likely exist, and will then test predictions

that ensue. However, even if this enterprise proves successful, it will necessarily capture

only a portion of the domains for which cultural learning is important. This is because

much of the information that must be acquired to succeed in any given society is

parochial in both content and type. For example, although it is true that public

performance is universally an avenue for achieving prestige, nevertheless, the nature of

such performance is so variable across cultures (and even across roles within a culture) as

to likely have made it impossible for natural selection to have provided substantial

foundations for the acquisition of the relevant information (e.g., while the category

“performance” may be a conceptual primitive, it probably does not subsume more

specific concepts, and thus lacks a rich structure that can serve as a scaffold for learning).

3.2 Domain-general cultural information acquisition mechanisms

Our species’ reliance on cultural information that is parochial in both content and type

suggests that natural selection may have favored the evolution of domain-general

cultural information acquisition mechanisms (DGCIAMs). At least two classes of such

postulated mechanisms are relevant to the present discussion; both address the problems

of the complexity and opacity of cultural information and related behavior discussed

earlier. First, selection may have favored the evolution of mechanisms dedicated to the

complementary tasks of pedagogy and the receipt of pedagogy. Second, whether the

agent serving as the source of cultural information is an active pedagog or a passive

26

target of imitation, in a crowded social world, learners must select whom to attend to as a

source of cultural information, a task that may be subserved by evolved mechanisms.

Csibra and Gergely (2006) argue that much cultural information transfer is

achieved through a goal-directed social process of teaching and learning contingent on

ostension, reference, and relevance. Ostension denotes the act of indicating that one’s

current actions are communicative efforts, thereby differentiating such behavior from the

stream of potentially-observable actions. Reference addresses the need to constrain the

topic of the communication from the class of all possible topics. Both ostension and

reference can be enacted by either the pedagog (who strives to indicate “now I am

teaching about X”) or the learner (who strives to indicate “I need information about X”).

Csibra and Gergely argue that specific cues, and the cognitive mechanisms that process

them, have evolved to facilitate, respectively, ostension (e.g., eye contact, eyebrow

flashing, and turn-taking contingency by both parties; motherese by pedagogs interacting

with infants) and reference (e.g., gaze directing / gaze-direction detection; pointing).

Relevance, a feature of the inferential process engaged in by the learner once a

teaching/learning interaction has been established, involves the presumption that actions

lacking an ulterior explanation are designed to convey information in light of the

learner’s state of knowledge.

To date, the work of Gergely, Csibra, and colleagues has focused primarily on

adult/infant interactions. However, ostension, reference, and relevance ought to

characterize all teaching/learning interactions, the only principal modification being that

bidirectional linguistic communication expands the channels available for ostension and

reference. A more substantial difference, however, between infant learners and older

27

learners is that infants have a vastly smaller range of potential pedagogs from which to

choose. Older learners therefore face the problem of selecting the targets from whom

they hope to learn. This is true both with regard to pedagogical interactions and with

regard to imitating a passive (non-teaching) model (while noting that the cognitive

mechanisms underlying imitation are deserving of attention from students of the evolved

psychology of culture acquisition, we leave this topic for another day—see Tomasello,

Kruger, & Ratner, 1993). However, the selection criteria differ for the two types of

targets. Because pedagogs scale their communication to the competence of the learner,

learners can usefully solicit pedagogy from individuals who possess vastly superior skills.

In contrast, because cultural information is complex and opaque, in the case of imitation

without pedagogy, learners must select targets that are closer to their own current

competence, else much of the model’s behavior will be subject to misinterpretation.

Differences between the two types of learning are reduced somewhat when the

issue of social structure is considered, as the distribution of cultural information over

roles is such that individuals will often benefit from seeking to learn from others whose

position in the social structure is not too distant from their own, since similar roles entail

similar resources, opportunities, and obligations (for seminal empirical work, see Harris,

1995). Lastly, the task of target selection is complicated by the bidirectionality of the

interaction: even imitation in the absence of pedagogy often has a bidirectional

component, as the target must tolerate the presence of the learner. Henrich and Gil-White

(2001) argue that, in exchange for access, learners grant higher status to individuals

whom they wish to imitate; market forces then influence individual choices, with less-

desired targets willing to grant more access. In contrast to Csibra and Gergely (2006),

28

who argue that the costs to the pedagog are such that pedagogy can be expected to be

primarily kin-based (as kin have an interest in the welfare of their relatives), we suggest

that the same market model applies to pedagogy: in both cases, the costs that accompany

being targeted by a learner can be outweighed by the elevated prestige and power that

flows from social support. The learner must thus trade off the value of the knowledge

possessed by the target (skill, success, etc.) against the costs of access, keeping in mind

that other learners are also competing for access.

The task of selecting targets from whom to learn is characterized by both

sufficient importance and sufficient overarching uniformity as to suggest that evolution

has created DGCIAMs for this purpose. We thus expect that actors will be adept at

identifying others who possess the optimal combination of superiority and role-relevant

knowledge; that observation of such individuals will be more acute, and subsequent

information better retained; that ostensive cues will be both displayed toward and sought

from such individuals; and that learners will be quite good at optimizing the access/cost

ratio in a fluid market.

An alternative strategy to selecting a single individual as the focus of learning is

to adopt the prevailing pattern of behavior in the local group. Because a single

individual’s success may be the product of many factors, raising the dual problems that a)

it may be unclear which aspects should be acquired by the learner, and b) some of these

factors may not be acquirable through learning, the conformist strategy will often prove

profitable (Boyd & Richerson, 1985; Henrich & Boyd, 1998). Note, however, that this is

not so much a method of learning as it is a method of deciding which of a number of

variants of behavior to adopt. This is because at least some understanding of those

29

variants (and hence some learning) must precede this decision; presumably, this process

must be iterated, with the actor coming to recognize finer distinctions among variants as

her command of the relevant information increases. Accordingly, many of the same

cognitive learning processes must underlie both information acquisition strategies that

target particular individuals as sources and conformist strategies that survey larger

numbers of individuals. There are also some parallels as regards the task demands of

target selection, since the relevance of behavior common across a set of prospective

models is in part contingent on their degree of similarity to the learner: behavior that is

common among individuals who occupy positions in the social structure similar to the

learner’s will generally be most relevant to, and hence should be most salient to, the

learner. We can thus expect some form of a DGCIAM to combine information about the

respective behaviors of actors with information about their relative social structural

similarity to the learner in order to efficiently promote conformist acquisition. Lastly, the

conformist strategy is less complex than individually-focused learning in regard to

calculating cost/benefit ratio in target selection: because the goal is to acquire the most

common variant of behavior, there is no shortage of potential models; hence it is a

buyers’ market, and the learner should be unwilling to pay much for access to prospective

models.

4 The effects of cognition on culture

Thus far, we have explored the extent to which cultural information shapes cognition, and

examined how the acquisition of cultural information may be underlain by evolved

psychological mechanisms of varying degrees of domain specificity. This may give the

30

impression that cultural information is a static feature of the environment. However,

because culture exists in the minds of individuals, the relationship between culture and

cognition is bidirectional, and thus dynamic. Specifically, because culture is instantiated

through processes of the transmission, retention, and application of information, the

composition of culture is subject to the influence of actors’ minds, as information that is

more likely to be transmitted, retained, and applied will come to predominate, while

information for which this is less true will become rarer, and may disappear entirely

(Sperber, 1996, Ch. 5).

4.1 Design features of mental mechanisms can influence cultural evolution

Anthropologists have long recognized that some ideas are “good to think” (Lévi-Strauss,

1962), meaning that they interdigitate with the mind in ways that make them attractive.

Originally, this notion was developed with regard to the manner in which the affordances

of real-world objects and entities (e.g., animals) facilitate symbolic distinctions that

usefully organize human social life. Such nebulous intuitions were later more rigorously

reconceptualized by cognitive anthropologists, who observed, for example, that folk

taxonomies tend to be structured in ways that complement features of short-term memory

(D’Andrade, 1995, p. 42-43). More generally, the relative learnability (ease of

acquisition, retention, and use) of cultural information can be expected to influence its

persistence and spread. With regard to information acquired via DGCIAMs, learnability

will in part be a function of the way that information is organized (as in the case of folk

taxonomies), and will in part be a function of features of social transmission (e.g., ceteris

paribus, because of the costs of pedagogy, ideas that can be acquired through imitation

31

alone will spread faster than those that require pedagogy). With regard to information

acquired via DSCIAMs, learnability will in part be a function of the extent to which ideas

contact the evolved content of the respective mechanisms (see also Boyer, 1998). For

example, Sperber and Hirschfeld (2004) and Barrett (2005) note that the special salience

of information regarding dangerous animals is such that ideas concerning dangerous

animals are more likely to spread and persist over time, even to the extent that erroneous

information about actual creatures (e.g., the belief that wolves often prey on humans, see

Sperber & Hirschfeld, 2004) and fantastical notions about nonexistent creatures

(Sasquatch, the Loch Ness Monster, see Barrett, 2005) become widely accepted (for a

pressing contemporary example, see Lombrozo, Shtulman, and Weisberg [2006] on the

obstacles to learnability of evolutionary theory). Similarly, using both experimental and

naturalistic data, Heath, Bell, and Sternberg (2001) have shown that the likelihood that

urban legends will be transmitted, and will persist, is in part contingent on the degree to

which they elicit disgust. Given that disgust is prototypically elicited by cues associated

with pathogen transmission (Curtis & Biran, 2001), this pattern is parsimoniously

understood as the result of the operation of a DSCIAM dedicated to the acquisition of

knowledge relevant to disease avoidance.

A general principle that can be expected to apply to many of the mechanisms

postulated thus far is that of error management (Haselton & Buss, 2000): natural selection

will have built biases into information processing systems such that, when these

mechanisms err, they do so “on the safe side,” making the mistake that, under ancestral

conditions, would have been the least costly of the possible errors. One consequence of

this is that, as illustrated by the above examples, DSCIAMs that address possible hazards

32

will lead to credulity, as actors will accept without evidence, and will not seek to test,

socially-transmitted information; aggregated over many individuals, the result will often

be a proliferation of wholly imaginary dangers. Consistent with this perspective,

preliminary evidence from an extensive evaluation of ethnographic materials suggests

that, around the globe, supernatural beliefs tend to depict a world filled with anxiety-

provoking dangerous agents and processes (Navarrete & Fessler, n.d.). At a more general

level still, the content of beliefs that describe hazards in the world will influence the

likelihood that those beliefs will persist over time. For example, many proscriptive

beliefs include notions of supernatural sanctions. We can expect that the most successful

proscriptions will involve sanctions that either a) are vague, referring to a wide class of

possible events (e.g., misfortune, etc.), or b) though specific, nonetheless refer to a

negative event that is relatively common, and for which the objective causes are not

readily evident (e.g., infant death in small-scale societies). Given the issue of credulity

mentioned earlier, proscriptions enforced by vague supernatural sanctions will be likely

to persist because bad things eventually happen to everyone; hence, when a taboo is

violated, it is inevitable that sooner or later events will unfold that can be interpreted as

evidence of the veracity of the belief. Likewise, proscriptions enforced by more specific

sanctions will be likely to persist if the events described therein happen with sufficient

frequency that, eventually, they will occur in the lives of rule-breakers.

4.2 Misfirings of other mental mechanisms can influence cultural evolution

In addition to the question of how design features of DSCIAMs and DGCIAMs influence

cultural evolution, a growing literature seeks to explain widespread cultural traits in terms

33

of the accidental misfiring of evolved cognitive mechanisms that are not dedicated to the

acquisition of cultural information per se. This approach arguably began with

Westermarck (1891), who hypothesized that incest taboos result from the accidental

triggering by third-party behavior of mechanisms that evolved to reduce inbreeding

among close kin. More recently, Boyer (2001) argues that beliefs in supernatural agents

are more likely to persist if they are minimally counterintuitive, meaning that the agents

possess nearly all, but not all, of the properties expected by mechanisms that serve to

detect agents and predict their actions; Fessler and Navarrete (2003) suggest that the

centrality of meat in food taboos is an accident of the salience of meat as a stimulus in

toxin-detection and pathogen-avoidance mechanisms; and Boyer and Lienard (2006)

argue that the combination of constraints on working memory, the operation of

mechanisms devoted to avoiding hazards, and the parsing of actions generates a non-

functional attraction to ritualized behavior. While we feel that efforts such as these are to

be encouraged, such hypotheses are necessarily developed on an ad hoc basis, in contrast

to the bottom-up predictive potential of a dedicated effort to explore the nature, and

consequences for cultural content and evolution, of mechanisms dedicated to the

acquisition of cultural information.

Conclusion

There are many ways to study the relations between culture and cognition. In this chapter,

we have discussed two traditional approaches in psychology and anthropology: searching

for psychological universals and for cross-cultural psychological differences.

Furthermore, we have attempted to describe what we take to be two of the most exciting

34

contemporary approaches to studying these relations: looking for domain-specific and

domain-general cultural information acquisition mechanisms, and identifying the

influences of the structure of our minds on how information is retained and transmitted.

35

References

Astuti, R., Solomon, G. A. and Carey, S. (2004). Constraints of Conceptual

Development: A Case Study of the Acquisition of Folkbiological and Folksociological

Knowledge in Madagascar. Oxford: Blackwell Publishing.

Atran, S. (1990). Cognitive Foundations of Natural History: Towards an Anthropology of

Science. Cambridge: Cambridge University Press.

Atran, S. (1998). Folkbiology and the anthropology of science: Cognitive universals and

cultural particulars. Behavioral and Brain Sciences 21: 547-569.

Atran, S., Medin, D. L. and Ross, N. O. (2005). The cultural mind: Environmental

decision making and cultural modeling within and across populations. Psychological

Review 112: 744-776.

Atran, S., Medin, D. L., Lynch, E., Vapnarsky, V., Ucan Ek’ and Sousa, P. (2001).

Folkbiology doesn’t come from folkpsychology: Evidence from Yukatec Maya in cross-

cultural perspective. Journal of Cognition and Culture 1: 4-42.

Avis, J. and Harris, P. L. (1991). Belief-desire reasoning among Baka children: Evidence

for a universal conception of mind. Child Development 62: 460-467.

Barrett, H. C. (2005). Adaptations to predators and prey. In D. M. Buss (Ed.). The

handbook of evolutionary psychology, pp. 200-223. New York: Wiley.

Berlin, B. (1992). Ethnobiological Classification. Princeton, NJ: Princeton University

Press.

Berlin, B. and Kay, P. (1969). Basic Color Terms: Their Universality and Evolution.

Berkeley: University of California Press.

36

Berlin, B., Breedlove, D. and Raven, P. (1973). General principles of classification and

nomenclature in folk biology. American Anthropologist 74: 214-242.

Boyd, R. and Richerson, P. (1985). Culture and the Evolutionary Process. Chicago:

University of Chicago Press.

Boyer, P. (1998). Cognitive tracks of cultural inheritance: How evolved intuitive

ontology governs cultural transmission. American Anthropologist 100: 876-889.

Boyer, P. (2001). Religion Explained, The Evolutionary Origins of Religious Thought.

New York: Basic Books.

Boyer, P. and Barrett, H. C. (2005). Domain specificity and intuitive ontology. In D. M.

Buss (Ed.), Handbook of Evolutionary Psychology, pp. 96-188. New York: Wiley.

Boyer, P. and Lienard, P. (2006). Why ritualized behavior? Precaution Systems and

action parsing in developmental, pathological and cultural rituals. Behavioral and Brain

Sciences 29(6): 595-650.

Brown, D. E. (1991) Human Universals. New York: McGraw-Hill.

Chomsky, N. (1986). Knowledge of Language: Its Nature, Origin, and Use. New York:

Praeger.

Chua, H.F., Boland, J.E. and Nisbett, R.E. (2005). Cultural variation in eye movements

during scene perception. Proceedings of the National Academy of Sciences, USA 102:

12629-12633.

Clark, A. (1997). Being There: Putting Brain, Body, and World Together Again.

Cambridge, MA: MIT Press.

Coley, J. D., Medin, D. L. and Atran, S. (1997). Does rank have its privilege? Inductive

inferences within folkbiological taxonomies. Cognition 63: 73-112.

37

Csibra, G. and Gergely, G. (2006). Social learning and social cognition: The case for

pedagogy. In Y. Munakata and M. H. Johnson (Eds.), Processes of Change in Brain and

Cognitive Development. Attention and Performance XXI, pp. 249-274. Oxford: Oxford

University Press.

Curtis, V. and Biran, A. (2001). Dirt, disgust, and disease: is hygiene in our genes?

Perspectives in Biology and Medicine 44: 17-31.

D’Andrade, R. (1995). The Development of Cognitive Anthropology. Cambridge:

Cambridge University Press.

Darwin, C. (1872). The Expressions of Emotions in Man and Animals. 1st ed. New York:

Philosophical Library.

Davidoff, J., Davies, I. and Roberson, D. (1999). Colour categories in a stone-age tribe.

Nature 398: 203-204.

Defeyter, M. A. and German, T. P. (2003). Acquiring an understanding of design:

Evidence from children’s insight problem solving. Cognition 89: 133-155.

Ekman, P. and Friesen, W. (1971). Constants across cultures in the face and emotion.

Journal of Personality and Social Psychology 17: 124-129.

Fessler, D. M. T. (2004). Shame in two cultures: Implications for evolutionary

approaches. Journal of Cognition and Culture 4(2):207-262

Fessler, D. M. T. (2006). A burning desire: Steps toward an evolutionary psychology of

fire learning. Journal of Cognition and Culture 6(3-4): 429-451.

Fessler, D. M. T. and Navarrete, C.D. (2003). Meat is good to taboo: Dietary

proscriptions as a product of the interaction of psychological mechanisms and social

processes. Journal of Cognition and Culture 3(1): 1-40.

38

Fodor, J. A. (1980). Methodological solipsism considered as a research strategy in

cognitive psychology. Behavioral and Brain Sciences 3: 63-110.

Galef, B. G. J. and Giraldeau, L.-A. (2001). Social influences on foraging in vertebrates:

Causal mechanisms and adaptive functions. Animal Behaviour 61(1): 3-15.

Gallistel, C. R. (2002). Language and spatial frames of reference. Trends in Cognitive

Sciences 6: 321-322.

Gelman, S. A. (2003). The Essential Child: Origins of Essentialism in Everyday Thought.

New York: Oxford University Press.

Gelman, S. A. and Wellman, H. M. (1991). Insides and essences: early understandings of

the nonobvious. Cognition 38: 213-244.

Gentner, D. and Goldin-Meadow, S. (2003) Language in mind: Advances in the study of

language and thought. Cambridge, MA: MIT Press.

German, T. P. and Barrett, H. C. (2005). Functional fixedness in a technologically sparse

culture. Psychological Science 16: 1-5.

Gil-White, F. J. (2001). Are ethnic groups biological “species” to the human brain?

Current Anthropology 42: 515-554.

Gleitman, L. and Papafragou, A. (2005). Language and thought. In K. H. and R.

Morrison (Ed.), Cambridge Handbook of Thinking and Reasoning, pp. 633-661.

Cambridge: Cambridge University Press.

Gordon, P. (2004). Numerical cognition without words: Evidence from Amazonia.

Science 306: 496-499.

Haidt, J. and Joseph, C. (2004). Intuitive ethics: How innately prepared intuitions

generate culturally variable virtues. Daedalus 133: 55-66.

Harris, J. R. (1995) Where is the child's environment? A group socialization theory of

39

development. Psychological Review 102(3): 458-489.

Haselton, M. and Buss, D. M. (2000). Error management theory: A new perspective on

biases in cross-sex mind reading. Journal of Personality and Social Psychology 78(1):

81-91.

Hauser, M. D. and Spelke, E. S. (2004). Evolutionary and developmental foundations of

human knowledge: a case study of mathematics. In M. Gazzaniga (Ed.), The Cognitive

Neurosciences (3rd Ed.), pp. 853-864. Cambridge: MIT Press.

Heath, C., Bell, C. and Sternberg, E. (2001). Emotional selection in memes: The case of

urban legends. Journal of Personality and Social Psychology 81(6): 1028-1041.

Heider, E. R. (1972). Universals in color naming and memory. Journal of Experimental

Psychology 93: 10-20.

Heider, E. R. and Oliver, C. C. (1972). The structure of the color space in naming and

memory for two languages. Cognitive Psychology 3: 337-354.

Henrich, J. and Boyd, R. (1998). The evolution of conformist transmissionand between-

group differences. Evolution and Human Behavior 19: 215-242.

Henrich, J. and Gil-White, F. J. (2001). The evolution of prestige: Freely conferred status

as a mechanism for enhancing the benefits of cultural transmission. Evolution and

Human Behavior 22: 1-32.

Henrich, J. and McElreath, R. (2003). The evolution of cultural evolution. Evolutionary

Anthropology 12(3): 123-135.

Hirschfeld, L. A. (1996). Race in the Making: Cognition, Culture, and the Child’s

Construction of Human Kinds. Cambridge, MA: MIT Press.

Hirschfeld, L. A. (2001). On a folk theory of society: Children, evolution, and mental

representations of social groups. Personality and Social Psychology Review 5(2): 107-

40

117.

Hull, C. L. (1943). Principles of Behavior. New York: Appleton-Century-Crofts.

Hutchins, E. (1995). Cognition in the Wild. Cambridge, MA: The MIT Press.

Ji, L., Peng, K. and Nisbett, R. E. (2000). Culture, control, and perception of relationships

in the environment. Journal of Personality and Social Psychology 78: 943-955.

Kay, P. and Kempton, W. (1984). What is the Sapir-Whorf Hypothesis? American

Anthropologist 86: 65-79.

Kay, P. and Regier, T. (2006). Language, thought and color: recent developments. Trends

in Cognitive Sciences 10(2): 51-53.

Kay, P. and Regier, T. (2007). Color naming universals: The case of Berinmo. Cognition

102: 289-298.

Keil, F. (1989). Concepts, Kinds, and Cognitive Development. Cambridge, MA: MIT

Press.

Kitayama, S. and Cohen, D., Eds., (2007). Handbook of Cultural Psychology. New York:

Guilford Publications.

Chiu, C.-Y., Leung, A. K.-Y. and Kwan, L. (2007). Language, cognition, and culture:

The Whorfian Hypothesis and beyond. In S. Kitayama and D. Cohen (Eds.), Handbook of

Cultural Psychology, pp. 668-689. New York: Guilford.

Levinson, S. C. (2003). Space in Language and Cognition: Explorations in Cognitive

Diversity. Cambridge: Cambridge University Press.

Levinson, S. C., Kita, S., Haun, D. B. M. and Rasch, B. H. (2002). Returning the tables:

Language affects spatial reasoning. Cognition 84: 155-188.

Lévi-Strauss, C. (1962). Totemism (R. Needham, trans.) Chicago, University of Chicago

41

Press.

Li, P. and Gleitman, L. (2002). Turning the tables: Language and spatial reasoning.

Cognition 83: 265-294.

Li, P., Abarbanell, L. and Papafragou, A. (2005). Spatial reasoning skills in Tenejapan

Mayans, Proceedings from the 27th Annual Meeting of the Cognitive Science Society.

Hillsdale, NJ: Erlbaum.

Lillard, A. (1998). Ethnopsychologies: Cultural variations in theories of mind.

Psychological Bulletin 123(1): 3-32.

Lombrozo, T., Shtulman, A. and Weisberg, M. (2006). The Intelligent Design

controversy: Lessons from psychology and education. Trends in Cognitive Sciences

10(2): 56-57.

Lucy, J. A. and Shweder, R. A. (1979). Whorf and his critics: Linguistic and

nonlinguistic influences on color memory. American Anthropologist 81: 581-605.

Machery, E. and Faucher, L. (2005). Why do we think racially? In H. Cohen and C.

Lefebvre (Eds.), Handbook of Categorization in Cognitive Science, pp

1009-1033. Amsterdam: Elsevier.

Mallon, R. and Stich, S. P. (2000). The odd couple: The compatibility of social

construction and evolutionary psychology. Philosophy of Science 67: 133-154.

Medin, D. L. and Atran, S. (2004). The native mind: Biological categorization and

reasoning in development and across cultures. Psychological Review 111: 960-983.

Medin, D. L and Ortony, A. (1989). Psychological essentialism. In S. Vosniadou and A.

Ortony (Ed.), Similarity and Anological Reasoning, pp. 179-196. New York: Cambridge

University Press.

42

Medin, D. L. Unsworth, S. J., and Hirschfeld, L. (2007). Culture, categorization and

reasoning. In S. Kitayama and D. Cohen (Eds.), Handbook of Cultural Psychology, pp.

615-644. New York: Guilford.

Navarrete, C. D. and Fessler, D. M. T. (n.d.). Does belief in the supernatural assuage

existential dread? An ethnographic analysis. Manuscript in preparation.

Nisbett, R. E. (2003). The Geography of Thought. New York: Free Press.

Nisbett, R. E. and Miyamoto, Y. (2005). The influence of culture: Holistic versus analytic

perception. Trends in Cognitive Sciences 9(10): 467-473.

Nisbett, R. E., Peng, K., Choi, I. and Norenzayan, A. (2001). Culture and systems of

thought: Holistic vs. analytic cognition. Psychological Review 108: 291-310.

Norenzayan, A. and Heine, S. J. (2005). Psychological universals: What are they and how

can we know? Psychological Bulletin 135: 763-784.

Norenzayan, A., Choi, I. and Peng, K. (2007). Cognition and perception. In S. Kitayama

and D. Cohen (Eds.), Handbook of Cultural Psychology, pp. 569-594. New York:

Guilford Publications.

Norenzayan, A., Smith, E. E., Kim, B. and Nisbett, R. E. (2002). Cultural preferences for

formal versus intuitive reasoning. Cognitive Science 26: 653-684.

Öhman, A. and Mineka, S. (2003). The Malicious Serpent: Snakes as a prototypical

stimulus for an evolved module of fear. Current Directions in Psychological Science

12(1): 5-9.

Piazza, M. and Dehaene, S. (2004). From number neurons to mental arithmetic: The

cognitive neuroscience of number sense. In M. Gazzaniga (Ed.), The Cognitive

Neurosciences (3rd Ed.), pp. 865-876. Cambridge: MIT Press.

43

Pederson, E., Danziger, E., Wilkins, D., Levinson, S., Kita, S. and Senft, G. (1998).

Semantic topology and spatial conceptualization. Language 74(3): 557-589.

Pica, P., Lemer, C., Izard, V. and Dehaene, S. (2004). Exact and approximate arithmetic

in an Amazonian Indigene group. Science 306: 499-503.

Regier, T., Kay, P. and Cook, R. S. (2005). Focal colors are universal after all.

Proceedings of the National Academy of Sciences 102(23): 8386-8391.

Roberson, D. and Davidoff, J. (2000).The categorical perception of colors and facial

expressions: the effect of verbal interference. Memory and Cognition 8(6): 977-986.

Roberson, D., Davies, I. and Davidoff, J. (2000). Color categories are not universal:

Replications and new evidence from a stone-age culture. Journal of Experimental

Psychology: General 129: 369-398.

Roberson, D., Davidoff, J., Davies, I. and Shapiro, L. (2005). Color categories in Himba:

Evidence for the cultural relativity hypothesis. Cognitive Psychology 50: 378-411.

Sousa, P., Atran, S. and Medin, D. L. (2002). Essentialism and folkbiology: Further

evidence from Brazil. Journal of Cognition and Culture 2: 195-223.

Sperber, D. (1996). Explaining Culture: A Naturalistic Approach. Cambridge, MA:

Blackwell.

Sperber, D. and Hirschfeld, L. A. (2004) The cognitive foundations of cultural stability

and diversity. Trends in Cognitive Sciences 8(1): 40-46.

Tomasello, M., Kruger, A. C. and Ratner, H. H. (1993). Cultural learning. Behavioral

and Brain Sciences 16: 495-552.

44

Tooby, J. and Cosmides, L. (1992). The psychological foundations of culture. In J.H.

Barkow, L. Cosmides, and J. Tooby (Eds.), The Adapted Mind: Evolutionary Psychology

and the Generation of Culture, pp. 19-136. New York: Oxford University Press.

Vinden, P. G. (1996). Junin Quechua’s children’s understanding of mind. Child

Development 67: 1707-1716.

Vygotsky, L. S. (1986). Thought and Language. Cambridge, MA: MIT Press.

Waddington, C. H. (1940). Organisers and Genes. Cambridge: Cambridge University

Press.

Wellman, H. M., Cross, D. and Watson, J. (2001). Meta-analysis of theory-of-mind

development: The truth about false belief. Child Development 72(3): 655-684.

Westermarck, E. (1891). The History of Human Marriage. New York: Macmillan and

Co.

Wimsatt, W. C. (1986). Developmental constraints, generative entrenchment, and the

innate-acquired distinction. In W. Bechtel (Ed.), Integrating Scientific Disciplines, pp.

185-208. Dordrecht: Martinus-Nijhoff.

Whorf, B.L. (1956). Language, Thought, and Reality. Cambridge, MA: MIT Press.

Wierzbicka, A. (1992). Semantics, Culture, and Cognition: Universal Human Concepts

in Culture-Specific Configurations. Oxford: Oxford University Press.

Winawer, J., Witthoft, N., Frank, M. C., Wu, L., Wade, A. R. and Boroditsky, L. (2007).

Russian blues reveal effects of language on color discrimination. Proceedings of the

National Academy of Sciences 104(19): 7780-7785.

45

i The two authors are listed alphabetically and contributed equally to this chapter.

ii These studies also show that cultural counting systems dramatically affect numerical

cognition. People with a limited counting system (a few number terms) are unable to

count beyond three and cannot accurately conduct arithmetic operations.

iii By contrast, as will be discussed at length later, selection favors developmental

pathways that rely on environmental inputs when the adaptive value of variants is

contingent on features of the environment that are stable within an organism’s lifespan

and home range, yet variable over larger expanses of time and space occupied by the

species.

iv For a different research program on culture and reasoning, see Medin and Atran, 2004;

Atran, Medin, and Ross, 2005; Medin et al., 2007.